Estimate fusion for 2D search sensors
Nassib Nabaa, Robert H. Bishop · Guidance, Navigation and Control Conference · 1995
In many realistic aircraft tracking scenarios, the quality of the state estimation can be reduced due to a poor geometry and insufficient data. This report presents a solution to the problem by considering an optimal fusion of the estimates of two Kalman filters using the same propagation model but diierent sets of search sensors measurements of the same target. For a typical engagement scenario, generated using a coordinated turn aircraft maneuver model, two extended Kalman filters are designed based on the Singer aircraft maneuver model. The filters operate from two distinct locations in space. The two nonlinear search sensors measurements available to the filters include range and azimuth, at a rate of 1 Hz. Simulation results show that, when both Kalman filters are performing optimally, the fused state e s timate enhances the accuracy of the estimation, e s pecially in the vertical direction, and therefore potentially allows a decrease in the reaction time.